September 10, 2024 • 46 mins
James Mackey, CEO of a leading RPO provider, SecureVision, and Elijah Elkins, CEO of Avoda, a highly rated global recruiting firm, co-host Workable’s CEO, Nikos Moraitakis, in our special series on AI for Hiring.
Nikos outlines the company's evolution from an ATS to a full-scale HR software solution. Discover how AI-driven tools are reshaping hiring processes—from resume screening to interview scheduling—empowering recruiters to work smarter.
Thank you to our sponsor, SecureVision, for making this show possible!
Our host James Mackey
Follow us:
https://www.linkedin.com/company/82436841/
#1 Rated Embedded Recruitment Firm on G2!
https://www.g2.com/products/securevision/reviews
Thanks for listening!
Mark as Played
Transcript
Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
Speaker 1 (00:00):
Welcome to the Breakthrough Hiring Show.
I'm your host, james Mackey.
Today, we are joined by myco-host, elijah Elkins.
Elijah, what's up?
How are you doing?
Hey, I'm all good, good, andour guest today is Nikos
Moriartakis.
Nikos is the founder and CEO ofWorkable and he's going to be
here today to talk to us aboutWorkable, telling us more
(00:21):
insights into the primary valueproposition, how they're serving
customers, and we're going tobe dialing in on AI use cases,
specifically about how Workableis thinking about AI and
feedback they're hearing fromcustomers, and what Nikos thinks
about the recruiting technologyspace right now in the
direction we're heading.
So, nikos, thanks so much forjoining us today.
How are you doing?
Speaker 2 (00:43):
I'm doing well.
Thanks for having me.
It's a pleasure to be here.
Speaker 1 (00:47):
Yeah, Nikos, why don't you just start off telling
us a little bit about yourselfand what you're doing over at
Workable right now?
Speaker 2 (00:53):
Workable is an HR software vendor.
Primarily we cater to small andmedium-sized businesses up to
maybe a few thousand employees.
We've been around for over adecade.
Everybody knows us forrecruiting because that's where
we started, although in the lastfew years we've expanded and
now we do all things HR.
But we've served just to giveyou a sense of proportion, over
(01:16):
30,000 customers over the yearsand we've seen 400 million
candidates for jobs, severalmillion hires.
So it's been around for a while.
We can talk more, a little bitabout what's going to be in 400
million candidates for jobs,several million hires.
So it's been around for a while.
We can talk more, a little bitabout what's going to be in the
future and what we're building,what the customers are doing.
But yeah, we're how can I saymid-size software vendor for HR,
mostly in Europe and the UnitedStates.
Speaker 1 (01:37):
Workable?
Did your team start in the HRISside or the applicant tracking
system side?
Where was the initial product?
And then how did you build uponthat and expand?
Speaker 2 (01:47):
Initially it was a straightforward applicant
tracking system.
We started 12 years ago.
Back then, the main propositionof SaaS software is I'm going
to give you better UIs,prioritize the user, give you
better experience and also maketechnology that previously was
(02:07):
available to the big enterpriseaccessible from price and
utility and ease of use tosmaller companies.
So Workable.
Started like that, like manyother companies of that
generation.
Over time, in recruiting, yourealize that one of the most
important things is sourcing isto be able to build a target
pipeline.
So we spent a lot of timebuilding one of the top partners
(02:31):
for Indeed and one of the toppartners for LinkedIn.
We move around millions ofcandidates every year.
We have our own sourcing tool,so we made an investment in that
and over the last three yearswe've expanded into HR
management, employee profiles,documents, time off, that sort
of thing.
Speaker 1 (02:52):
Yeah, I like the strategy and so it sounds like
you.
When did you expand intosourcing?
I'm just curious from atimeline perspective.
Speaker 2 (03:00):
It was around 15, 16, when we put a lot of emphasis
in sourcing.
Oh, that's awesome, and alsothat happens to be the genesis
of all our AI technology thatwe're going to talk later,
because when you go intosourcing, it's one of the places
where this is even earlier and,let's say, less strong
technology than the one we haveright now, where applicable.
Speaker 1 (03:20):
That's great couple.
That's great because the reasonI asked that question is I'm
curious.
My assumption is that yourcompany weathered the storm of
market consolidation a littlebit stronger than some of your
competitors, due to the factthat you built out a product
suite more so proactively, priorto COVID and the market crash
in 2022.
I've spoken with a lot ofrecruiting technology leaders
(03:42):
that are essentially scramblingto develop a product suite to
make their products stickierright, like maybe they were just
only doing sourcing and nowthey're scrambling like, okay,
we really need an ATS, becausecompanies aren't they're not
going to have severaltechnologies in their stack,
particularly if they're SMB andmid, lower, mid-market focused.
And so then I talked to otherslike who are ATS CEOs that are
(04:04):
like, okay, we really want toget into potentially more of the
onboarding HRS side or we wantto get more into the sourcing
side, and they're essentiallyjust trying to make decisions
based around, likeactively, tryto expand your product suite,
and probably something that Iwould again, this is a total
assumption.
I don't know, but I would say,compared to other recruiting
(04:31):
tech companies, probably enabledyou to have a little bit more
consistency and minimize churnas much as possible, which, of
course, I'm sure for everybodyit's still increased, but
weathered the storm a little bitmore than some of the other
CEOs that I've spoken with overthe past few years.
Speaker 2 (04:46):
Look, you touched on something a little bit more
fundamental about how the marketmoves.
I think we are at the end of acycle of innovation, so usually
you have a decade.
There was a big decade for B2BSaaS software.
It became easier and easier tomake it.
Players played well with eachother, there was a lot of
financing, there was a lot ofappetite in the market.
(05:08):
The market was going up,companies were willing to spend
for point solution, for a smalledge on competitiveness and
there was a lot to be gainedthere and this resulted in a
proliferation of a lot ofsoftware and a lot of software
for very, very specific usecases, from point solution down
(05:31):
to what would be features of apoint solution.
Now, obviously, this clears upat some point.
So naturally, the market isgoing to consolidate.
I think it's driven more bothfrom the industry, as you said.
For each one of those companiesit's a stronger play to be able
to expand into more things, tosell more things, to grow their
footprint on the customer, tomaybe consolidate through M&A
(05:52):
and wicked by the financing withothers and, on the other hand,
the customers as well.
I don't know about you guys,but I hear the story very often.
I am the new HR manager.
Previously, the company wasgrowing fast and somebody bought
expensive software foreverything.
Now cleaning up this mess, Iwant an affordable solution that
(06:13):
does everything.
I bet it's happening everywhereit is.
Now, in our case specifically, Ithink we got a bit lucky.
We were one of the first tocome into this space, so we
built some scale.
We already had numbers that wewere a profitable company with
(06:33):
strong economics.
So, even though the 22 and 23were completely terrible years,
it wasn't a huge setback for us.
And we were lucky as wellbecause, for our own reasons, we
ended up becoming profitableright at the year of COVID,
before we knew that this wasgoing to happen.
(06:54):
So we didn't have to retraceour steps like many companies
had to do during these years,and we said since we're not
going to sell a lot over thenext two or three years, we
might as well make a new productto improve our retention.
Speaker 1 (07:07):
That's pretty smart.
Yeah, lucky or not, I'm reallyhappy for you guys.
It's a strategy that seems toit just makes sense.
It makes a lot of sense andyour team was able to pull it
off, it seems, a few yearsbefore a lot of recruiting tech
companies.
I know recruiting techcompanies are just rolling out
additional products.
In the past year or so theyprobably already churned a lot
(07:28):
of customers that they might'vebeen able to hold onto if they
had done this proactively.
Speaker 2 (07:34):
I think through M&A or product expansion, you're
going to see this market startconsolidating a little bit over
the next few years.
Speaker 1 (07:41):
Yeah, for sure, for sure, Elijah, you have any
follow-up thoughts on that topicfor?
Speaker 3 (07:45):
years.
Yeah, for sure, for sure,elijah, you have any follow-up
thoughts on that topic?
I've actually implementedWorkable a couple of times at a
few different companies.
I didn't mention that to Nikosin the beginning, but so very
familiar with the product.
I haven't used it for a coupleof years but it was super
interesting to see you all hadsourcing before anyone else did,
before the other ATSs, beforeAshby came out with kind of
(08:06):
their sourcing tool with thatkind of all-in-one workable was
had that TA product suite before.
I'm curious you don't see toomany companies right, start with
the ATS, have a very robustsolution and then build out an
HRIS or an HRMS system.
You've seen HiBob BambooHRRippling right try to add like
(08:27):
recruiting modules, like anATS-like functionality to their
HRIS.
What does that kind of looklike for you?
Is the ATS almost a standaloneproduct side by side and you
switch from one user experienceinto the other, or is it
holistically integrated to oneor just yeah?
How does that kind of look tobuild out an HRIS at workable?
Speaker 2 (08:47):
that's a very good question.
It's a different problem whichdirection we go to, and a lot
can be said about this, but letme try to focus on a couple of
interesting bits.
I think what we're doingcommercially is more difficult,
because you may have a bunch ofcustomers who are even excited
(09:08):
about your product but they'renot the buyers of the other
product.
So you're typically not.
You may be talking to the HRdepartment but you're not
talking to the same buyer.
So essentially, commercially,you need to rebuild a lot of the
relationships and marketing andbrand that you need to build.
And it's also usually therecruiting is a sub-department
(09:31):
of whoever buys the HRIS.
So nobody's going to buysoftware for their boss or for
some of the people who havesomething very separate.
So expanding in that manner isnot tremendously easy.
On the other hand, you haveanother force into it that most
companies, as I said, wouldrealize.
It's better to have onesoftware where everything from
(09:53):
the org chart down to theterminations, to the backfields,
to the acquisitions, where itunderstands my people, it
understands who performs well.
Now, with AI, you can believethat understanding performance
is going to link to betterrecruitment.
So when customers go for thisbecomes a question what is
better to buy if you assume,let's say for simplicity, that
(10:16):
the customer says workables HRISis the newest weakest and
bamboos recruiting is the newestweakest, which one do you hurt
more by accepting a lower end?
Speaker 3 (10:27):
solution.
Speaker 2 (10:28):
The argument there, and my hope is that recruiting
is the one that is a performanceapplication.
It does make a big differencein results.
The other one is a databasewith a good UI, let's be honest.
So would you rather have morecandidates or a different
signature?
The signatures are the same.
Speaker 1 (10:50):
When it comes to incorporating AI at Workable,
I'm curious to see howaggressive Workable is with
implementing AI functionality.
My assumption I think youtouched on this is are you
starting more on the sourcingside?
But I'd love to learn moreabout that.
But I think some of the moreestablished players I've seen
are moving maybe a little bitslower, not necessarily because
(11:14):
they have to, but just becausethey have a full product suite.
They're taking their time to bea little bit more thoughtful
about what they're rolling out.
I'm curious are you pushing AIaggressively or are you just
trying to roll it out in verysmall little phases that you're
more confident, or high impact?
What's your kind of AI strategy?
How do you think aboutincorporating AI into a product
(11:39):
or pushing out a feature or likewhere to start?
I'm just curious about, like,your thought process and then
that leading into where did youactually start?
Can you walk us through yourprocess?
Speaker 2 (11:50):
The short answer is that we are mega aggressive with
AI.
We have seen fantastic results.
We already have a ton of thingsin production available to all
our users, and thousands ofusers daily, of many different
things we do with AI.
We're extremely enthusiasticabout the results.
Yes, it's a new technology.
(12:12):
Yes, harnessing it and managingall this is not as simple as it
sounds, but I think everythingis going to move in that
direction and I thinkapplications are going to change
in shape as well.
This is going to change UIs.
It's going to change a lot ofthings, and there's absolutely
(12:32):
no way we're waiting on thisgame.
We want to take every problemthat there is in there and try
to crack it with AI dooms andcrack the code of how it's going
to be represented, interactedwith, et cetera, in the future.
Let me give you a sense ofwhere we are right now.
So today in Workable, we'llwrite your job descriptions with
(12:54):
AI, and already 30% of the jobscreated in Workable today are
written by AI.
This is not a small.
This is a big use case.
We publish 700,000 jobs a yearthrough our customers and a
third of them are already donewith AI.
We have salary calculators basedon our data.
(13:16):
Ai helps you identify whatsalary you need to do.
It creates interview kits, itcreates scorecards, it creates
video interviews for every job.
It functions like a juniorsourcer and it's going to go and
find you 50 passive candidateslike a recruiter would, at least
(13:37):
as a target who are matchingwith the job.
It's going to write the emails,but highly personalized, based
on their experience andeverything that he knows on the
internet about them, to recruitthem.
Those emails have twice theresponse rate of personalized
emails from recruiters andalmost four times the response
rate of templates.
(13:58):
And then, for all the candidatesthat come into the system,
anything that you put in thatyou see on your interface we
scan all of them.
For each job description, wehave an AI that essentially
determines what should be thecriteria for screening for that
job and then it marks every oneof those criteria.
So, basically, on top of everyprofile you see a little
(14:21):
scorecard what are the fivebullets you want and which ones
are green.
So it it does all the readingof the resumes and
interpretation and screening foryou, and all of these things
are live use cases like they'renow on the system.
You can go get a trial accountand use them, and thousands of
people are using them every day.
So there's an endlessdiscussion about where this is
(14:44):
going, how good it is, and allthat good stuff, but it's clear
to us that this is going to be aproductivity bomb.
Speaker 1 (14:53):
Okay, so we covered a lot of ground there too.
I actually had to write somestuff down because I want to
slow down on it.
I wasn't, honestly.
I thought you were just goingto tell me like, oh, we were
going to, we can write emailsfor your sourcing team, like a
lot of companies.
That's where they're at.
Oh, we generate a sourcingmessages for you.
Okay, we talked about, ofcourse, like the AI generated,
jd, which is, honestly, that's.
(15:14):
A lot of companies are doingthat one Interview kits,
scorecards.
Now for the interview kits.
Can you provide us a little bitmore detail on what that
includes?
Speaker 2 (15:26):
I'd like to dial into that a little bit more
Basically.
For an interview kit you arepretty much the best way to
describe it.
You're pretty much a structuredinterview.
Okay, you have some areas thatyou want to touch and for each
one of them you have a fewquestions and next to it, if you
have a scorecard, you also havea way to note how well you felt
(15:47):
that person responded to thatquestion on your notes.
Now, this is a great practice.
The problem is to construct thisset of questions and to have a
comprehensive interview thatanswers the right points and
have imaginative ideas forquestions.
That's hard.
That's what the AI does.
It basically writes that foryou.
So, let's say you were hiring afinance director, it will
(16:09):
create a section about.
It says it sees director.
It will create a section aboutleadership, managing teams, et
cetera, and have some questionsabout this.
But he will also ask questionsabout the specific jobs and if
you do a video interview, hewill create the video interview
questions for you.
So he will prepare the videointerview, what should be asked,
(16:30):
etc.
Which is a slightly shorterform version of a scorecard.
In fact, the thing.
Okay, this is new.
We have a way as well to forthis, to listen in on the
interview and transcribe thenotes.
But the funkiest part thatwe're going to do soon is that
essentially it's going to go tothe scorecards and evaluate the
(16:52):
candidates based on what itheard on the interview.
It's going to write theevaluations for you, so it
drafts it for you, you check it,that it's fine, you make any
corrections and you're done.
And then the other thing thatfor next year is that obviously,
if you're looking at a lot ofpeople, you would like to see an
inference report, somebody likea recruiter telling you we've
looked at these five candidates,this one was stronger.
(17:14):
They're going to write you abit of an essay where it
synthesizes for every area who'sthe strongest and et cetera.
So basically the other thingthat we're making now we have
shown a little video of a demoof it.
We haven't made it public yet.
We have an agent.
He has an email in a Slack.
(17:35):
You talk to it on Slack likeyou would with an employee and
essentially it goes and doesthese things I just said for you
.
So you say schedule the nextones for interviews, move them
along, tell me what was theresult of all the interviews,
start to make the job andbasically you talk to it like
you would an employee and itgoes and accomplishes tasks and
(17:56):
it also comes back to you andsay I got three emails from
those candidates.
One is not interested, I willreject him.
The other two are interested.
I prepared a response for youto review.
Click here and go on.
This is how you should imagineit in the future.
So anyway, I'm getting excitedwith these things and I just go
(18:17):
off on a tangent, so I'm sorry.
Speaker 1 (18:18):
Oh yeah, the tangents are the best part.
Let's just go down like therandom stops where we just go
down these random rabbit holes,which is like the whole point
here.
I think that's again the valueout of this podcast in general
is the hosts, as well as theguests, are experts in the space
, so we can get a little bitmore technical, which is, I
think, what we like to do andwhat our buyers and what
Workable's buyers they want tohear.
(18:39):
They want to understand thedifferentiators, how and even
it's not only just like yourcurrent functionality, it's like
how you think about AI, becausethat's going to inform your
product development and roadmap,right, if they feel like, okay,
this guy, nikos, really has asolid pulse on this stuff.
He's doing some innovativestuff in the space that we're
not hearing from otherwell-known recruiting tech
providers.
Like, maybe we should bet onthis guy and workable versus
(19:00):
some of the other ones.
So it's going down these rabbitholes is actually, I think,
really important.
Yeah, it's fun for us, first ofall, selfishly, but it's also
something that I think peopleneed to hear.
So I guess, like I and Elijahjust cut me off.
Man Like I'm, I have so manyquestions, questions, but I I
like one one thing real fast isnow you're talking about, like
the video interviews you talkedabout.
(19:21):
Essentially, are you guys alsodoing like the ai note taking,
where you're plugging intointerviews and it sounds like
you said you're going to startto put together evaluations of
the transcript.
Speaker 2 (19:31):
so you're doing that, okay.
But the interesting thing isthen when you read the
transcript Okay, so you're doingthat, okay.
But the interesting thing isthen when you read the
transcript and you alsointerpret it and help the
interpretation.
But that is for later.
Speaker 1 (19:44):
So for the interpretation piece like what's
interesting is that we've nowspoken with the CEO of
BrightHire, ceo of Pillar andthen a couple of CEOs of smaller
companies a little lessestablished in the space, and
what was interesting aboutPillar and BrightHire is neither
of them are actually providing.
It's like they're providingsummaries and they're packaging
(20:07):
the data in a consumable way sothat hiring managers can compare
and contrast candidates andthen also evaluate.
Okay, you have three stages ofinterviews, you've covered 75%
of the things you need to, butwe didn't ask this one question
in the screening call.
We need to double back on that.
Or they're providing some kindof have, we asked all the
evaluation criteria, but they'renot necessarily scoring
(20:28):
candidates.
It doesn't sound like theproducts are saying, hey, here
are your best candidates, orthey're not really doing an
evaluation piece or a scoringpiece.
We spoke to a CEO of a smallercompany called Qual and they're,
I think, a very early stagecompany.
They do scoring.
They do scoring.
Now it's different.
(20:49):
They're working with highvolume, light industrial workers
where apparently, he said, theresumes are typically very
incomplete so it's verydifficult to tell who's worth
speaking to, and so the scoringaspect is important because
there's thousands of applicantsand there has to be some kind of
way.
That's not doing a keywordmatch.
That's not effective.
So maybe because Bright IronPillar are more in the corporate
(21:12):
type of space, if you only havefive candidates in the
interview process, do you reallyneed a stack rank or like a
scoring mechanism in place?
I'm curious how you look atthat topic, like when you're
thinking about the note-takingaspect, packaging the analysis
essentially, or summarizing.
Do you see Workable going inthe direction where you're
(21:32):
actually going to be scoringcandidates and saying hey, or do
you see yourself staying awayfrom that?
And how important do you thinkit is?
Why would you do it or not?
What do you think?
Speaker 2 (21:42):
No, because I don't think that's what you need in
recruiting.
I think most people, especiallypeople outside the space,
imagine that recruiting is aprocess that involves a great
degree of judgment and rankingpeople and making some
evaluative assessment on them.
(22:04):
In reality, it's not so much.
It's more of a filteringprocess.
It's a process where you try toarrive at a few people for whom
you have collected theappropriate information and you
have verified that they are anappropriate fit.
Let me put it in simpler wordsIf you give me a short list of
five resumes of people who workthis job, have the right
(22:27):
qualifications and we haveverified a few things around
them, my job is easy.
That's not what I want thesoftware to do.
That's an easy task.
The computing part in recruitingcomes into the fact that you
have to process a lot of thingsfrom many sources and you have
to go through many steps.
It involves many channels, fromjob boards to emails, to this
(22:48):
to that.
So it's more of an informationorganization process, as you
suggested.
We see, as you said before,many vendors say I'm going to
organize the information for you, I'm going to present it to you
, transform it, process it.
Take something that was said inan interview, write it in words
, put it in the right box so youcan see it next to the right
(23:09):
things and make a decision.
The decision is neither themost important thing for the
software to do, and I think it'sthe thing that the software is
least qualified to do, becauseit's the most idiosyncratic as
well.
Think about it Honestly at theend of the day, people do have
companies, departments, teams,whatever it is.
(23:30):
They have different view ofwhat is important to them.
The point is to get them to thepoint where they have a
meaningful choice between a fewvery qualified candidates for
which they have done their duediligence, so to speak.
They've got the rightinformation.
When you hire an executiverecruiter and you pay top dollar
(23:52):
for that job, what do you get?
You get all the resumes, youget all the information, but you
also get some organized dataand comparative things around
them and you get a little bit ofa narrative.
Somebody, essentially, hasprepared a dossier with you and
has checked everything, and heknows that you could hire any
one of those people.
(24:12):
Just pick the one you like.
That's what the software shoulddo.
The other thing about scoringand ranking it's also that it's
touchy.
You can really go wrong there.
Do you want a world where anyvendor can make a tweak on the
algorithm and that changes whogets hired?
(24:32):
We don't want to get into that.
I don't know if I'm willing tobelieve that sometime in the
future AI might be better andmore objective than humans,
because humans are flawed forthings like that.
Maybe sometime in AI is goingto be better, but not today, not
next year.
Speaker 1 (24:55):
What about top of funnel?
So okay, not hiring decisionscoring and I don't mean scoring
necessarily like 98%, I justmeant okay, of the criteria that
you're looking for.
There's a group of folks thatseem to be more aligned with
that.
Whatever like loose, right?
Okay, so not.
Maybe not at the end of theprocess.
It's more packaging, the data,the narrative, as you put it,
organizing, I think, the slackaspect of what you talked about.
(25:16):
Hey, reach out to these nextfolks.
That's awesome.
I'm really excited to see thatproduct.
That's going to be so cool.
But what about top of funnel?
Right, so you get 1,000applicants?
Do you see value in any kind ofautomated screenings?
So AI can help package, saying,okay, out of these 1, these
thousand people, you'reessentially allowing them to
(25:38):
engage in a bit of the screeningprocess upfront so that you can
have a better way ofidentifying a short list than
just do a resume keywordmatching situation that, either
done by a person or anapplication, is not necessarily
the best way to figure out whoto actually jump on a call with,
like with a person.
So do you see any kind ofscoring benefit on top of funnel
(26:01):
?
Or still you're just you don'tlike it?
Curious there.
Speaker 2 (26:04):
Even on the top of the funnel, where indeed there
is more motivation, because youreally need to do a first filter
and, to be honest with you also, we're not the government.
We're not giving peoplebenefits to the extent that it's
reasonably objective, even ifthere was someone who was good
and you didn't manage to capturethem, it's survivable and
humans do it too.
However, I think it isextremely hard to create a
(26:31):
scoring mechanism that isn'tgoing to go wrong.
It's such a complex problemlike.
Even if you have five criteria,we all know they do not all
have the same weight.
The job description doesn'ttell it either.
It's just a document.
It's not in your head, so Idon't.
I would be extremely reluctantto make a system.
We could make a system likethat with Merka, but it would be
(26:53):
good many of the times.
It could probably beat some notvery good recruiters.
That's not what you need, whatyou really need.
In screening.
You have a big task.
You have to go through a lot ofresumes and you have to
impartially, objectively andconsistently evaluate the facts
(27:15):
on the resumes versus a list ofthings you're looking for.
Nobody really does this.
It would take you a few minutesper resume.
Nobody really does this.
In fact, most people, actualrecruiters what they do they
operate more like keywordmatches.
They look for the resume for afew things that seem to light a
bell on them and then they movethem forward.
And let me tell you thebusiness reality of the thing.
(27:38):
A few more resumes come a fewdays later.
Do you think they get the sameattention from the same person
with the same criteria?
So there's a very inconsistentprocess from getting to the big
amount of information that's abig chunk, which is all the
resumes to the summary of that,which is all these people on the
five things we need will do thesit.
That is what AI can do for you,and then AI doesn't need to be
(28:02):
involved in the decision becauseit has condensed the
information, does the legwork somuch for you that now it's easy
for you to just look at themand apply your own criteria on
top of that.
Speaker 1 (28:13):
It's like a data organization presentation.
Speaker 2 (28:17):
I wouldn't do scoring , because scoring suggests that
you take all that information toa formula and say this guy is a
10, is a 20, is a 50.
That is what does it mean.
It sounds silly even when yousay it.
What does this mean?
But what it can really do is toshow you a bunch of cards of
people with some informationthat makes you easy to make a
(28:40):
decision.
Let's think of it realistically.
You see, someone who meets allfive criteria, move them on,
someone in between?
Okay, I can still look andverify, but it really makes it
easy to pick up things that arecompletely off or completely on.
Which is the majority?
(29:02):
Okay, when you get the pipeline, the majority.
If you take the ones that arebang on and the ones that are
white in the sky, apply here youhave 80% of it.
Speaker 1 (29:11):
Yeah, that's okay, that's super helpful.
Thank you, elijah.
What do you got, man?
What questions you got in yourhead?
Speaker 3 (29:17):
Yeah, two quick questions.
One's a comment.
Thank you, elijah.
What do you got, man?
What questions you got in yourhead?
Yeah, two quick questions.
One's a comment, the other is aquestion.
Have you guys considered or areyou working on anything that
analyzes all of the scorecardsand the feedback and provides
almost like a state of the roleand kind of trends that seem to
be things going well withcandidates being passed through,
(29:37):
or things not going well,because I've never seen that,
but that's often right.
What a more strategic kind ofrecruiter is trying to figure
out is what are these trendswith the candidates being passed
through?
Why are candidates gettingrejected?
And you don't always have theright rejection reason
categorized appropriately in theATS.
It may actually be within dataor the scorecard notes or
(30:01):
something.
Does that make sense?
Speaker 2 (30:04):
Yes, but not with AI.
Workable allows you to do this,both to classify the reasons of
rejection and obviously whenyou get the scorecard.
There is a little bit more howcan I say?
Informal enforcement.
It's very hard for someone whomeets all five criteria to say
it wasn't good enough on thefirst screening.
(30:25):
So it keeps you honest a littlebit because it surfaces more
information more deliberately infront of you, but automatically
, no, we don't do it.
Now, the other thing you saidto try to even create a profile
for the role.
I haven't thought of it thisway.
What I think, what I would likeas a user, I would like the AI
(30:46):
to get to the point where, inone way or another, it tells me
remove this requirement.
It's stupid, you're notselecting for it.
You said five years ofexperience and you didn't honor
it.
Speaker 1 (30:58):
stop I wish I could talk to my customers that way as
a like secure vision bycompanies rpo provider.
I wish I could get on the phonewith a demanding ceo of a
growth stage sass company and belike remove this requirement.
It's stupid, I end up doing it,but it takes a lot more tact
and time and data.
To summarize, but if I couldhave a software do it for me,
(31:19):
that'd be great.
I could say, hey, it's allright here.
Speaker 2 (31:23):
What we are hoping at Workable and I'm not going to
say that we really do this inany meaningful way, okay, only
indirectly.
Is that eventually Workablewhen it writes the job
description which, as you said,to get a GPT or something
similar to write?
It is not very impressive onits own.
It's to the point where itstarts writing the job
(31:44):
descriptions for better successfor you knowing that stuff.
But right now you will seeapplications using AI technology
let's say, black boxes for themoment, using this new
technology, plugging it intospecific places and improving a
part of the experience and thenmaking it informed by more
(32:04):
information.
And when we collapse the dam ofinformation and start sharing
between for your account, whereone can inform its decisions
from things that happenelsewhere in the application
that you might see the reallybig steps change, where this
thing makes things better foryou because it has also observed
(32:26):
the way you work.
We're not there yet, but wewill be there in a few years.
Speaker 3 (32:34):
Is that part of the building?
The HRIS side of things isreally so you get that whole
picture of, like talentmanagement, who's being promoted
, who's at the company, theirperformance, and that can
actually feed back into therecruiting side of things.
Speaker 2 (32:50):
Yeah, let me tell you futuristic features I want.
When somebody quits to have aclone button and this thing
understands this employee, theirrole, who they work with.
It probably already knows theirpersonality tests etc.
And then goes out to source andrecruit people against that and
(33:11):
even tells you how theirpersonalities will fit Remove
the personality part, whichsounds a little bit dodgy.
Something that understands yourcompany, your needs, something
that gets your hiring plan andknows how much time it's going
to need to fulfill thosepositions, because it takes us I
don't know three months to geta batch of HDRs.
So I want all my orgs to beable to do my future planning
(33:33):
and my acquisition planning andmy succession planning.
And this thing understands thepeople.
I want a referee on theperformance review, somebody who
has gone into salesforce and toapplication and has read things
about the employee and knowsthe one-on-ones and tells me and
helps me do the performanceevaluation and compares people
(33:54):
who's the best in the departmentof this or that.
There's so many things you cando.
This is futuristic.
I'm not promising somethinglike that now, but it is totally
within the realm of thepossible and I think the most
interesting thing with AI is notgoing to be that it's going to
give like one absolutely wow upthat stuff.
Like that are going to startsounding.
Ai is going to get intoeverywhere and eventually smart
(34:17):
companies are going to startsounding.
Ai is going to get intoeverywhere and eventually smart
companies are going to createinteresting UIs that do these
things.
Imagine we're in 1997 and say Iwould like a place where it has
all the videos and all themovies in the world and I can
watch them by connecting to theinternet.
It would happen.
Speaker 1 (34:36):
Similar things would happen with that when we were
talking about some of theinitial ai applications, there
was a follow-up question I hadon sourcing.
But before I switch over tothat topic, elijah, do you have
any other questions you want todial in on here?
Speaker 3 (34:48):
I'm just curious quickly, nikos, how do you feel
like workable and other but?
what I wouldn't call legacy ATSsright, because it was only 10
years ago.
How do you feel like you'llcompete and what the dynamic
will look like?
Competing against more like AInative maybe you could say
(35:09):
recruitment software companiesor HR software companies that
are coming out of places like YCombinator today, where they're
starting with Gen AI day one andarchitecting everything around
that?
Do you think companies likeWorkable still have the
advantage because of theexpertise and knowledge and team
size and customers and all ofthat, or that they actually may
(35:31):
have an advantage in some ways?
I'm just curious.
Speaker 2 (35:36):
Look, any new company has the advantage that they're
starting from a blank slate.
Okay, this in some years, mostyears is not an advantage, it's
a liability when there's a newtechnology coming into place and
everybody needs to adapt.
What did I say in the beginning?
(35:56):
This is going to change UIs andhow we think about how this
happens and all that stuff.
Obviously, if there's a newoptimal state, there's a new
best design for that thing witha new technology, the one who
starts from a black slate has achance of getting there faster
than the one who has to redesignthings there faster than the
(36:16):
one who has to redesign things.
So the question becomes where isthe change going to happen?
Right now there is.
I don't want to strawman it,but there is a little bit of a
how can I say?
Simplistic view that, oh, sinceAI can do everything, we can
(36:37):
write the code.
We can do everything with AI.
Basically, we replace a lot ofthe business logic,
functionality, workflows,processes, right, yellow that
the software had with just an AImaking a decision or making it.
It's not good enough to getthat point.
You can't just get an AI andsay, oh, you be my recruiter and
now we have no UI.
I just talk to you and you doeverything.
(36:58):
We might get there one day, butI'm not even sure that would be
the ideal interface for a bigcompany.
But I do not fear that somebodyis going to use AI to replace
what a big business softwarelike Workable does.
I'm not afraid of that at all.
I think the technology is goingto come and slot in.
(37:18):
Just because you made a newplane, you can't throw away your
airports.
Somebody's going to post thisjoke to LinkedIn.
There's so many things thatneed to happen.
What I think will be disruptiveis the second wave of those
companies that come in withdifferent premises.
Sooner or later we will startto see an emerging pattern on
(37:40):
how new software is going to beorganized and new UIs, somebody
coming with a good intuition andvision about how that would
apply to the particular domain,like we did 12 years ago A
different design of how theapplication should be.
And they use AI technology toget there much faster.
(38:02):
And because they built thiswith the assumption it's going
to be done with AI, they are notincubates from the requirements
of our users, who would not bewilling to put up with such a
dramatic change, I'm guessingI'll tell you in five years.
All right, sounds good, butwe're not willing to put up with
such a dramatic change.
Speaker 3 (38:20):
I'm guessing I'll tell you in five years.
Speaker 2 (38:21):
All right, sounds good, but we're both heading in
the same direction, and that'sthe important thing.
Speaker 1 (38:25):
Good.
So I have one question aboutsourcing, and then I was hoping
I could ask you about yourLinkedIn profile photo.
We're eating the watermelon.
I want to make sure we haveplenty of time for that.
But when it comes to thesourcing functionality, we had
plenty of time for that.
But when it comes to thesourcing functionality, you're
saying it's like the AI can goout and pull candidates right,
like I think you said that right, the AI could say, hey, here's
some great candidates for you toconsider, right, okay.
(38:45):
So if that's the case, thenlike, where does it go?
Where is it pulling this datafrom finding the candidates?
How's it doing that?
Speaker 2 (38:54):
Earlier, I mentioned that for a few years we spent a
lot of time developing sourcingtechnology.
Essentially, what we've builtis through a network of
partnerships with job sites,with data providers, etc.
We've built a pretty decentdatabase for looking for talent.
It's not as big as LinkedIn.
(39:17):
It's not as accurate.
It's basically a composition ofmany sources.
However, for the purposes of anAI identifying people that
would be suitable or a goodmatch for the job, it's
perfectly good because it's muchyou're going for.
So, essentially, we have ourown proprietary database that
(39:37):
we've built over the years.
Just to be clear, this doesn'tinvolve our customers, right?
Yeah, obviously, I'm assumingyou guys understand that.
There's no way this could bethe same thing.
Speaker 3 (39:49):
Yeah, yeah, yeah, different data.
Data is completely separate.
Speaker 2 (39:54):
Obviously otherwise, if we could, if the data of our
customers was our own which itisn't it would be one of the
biggest in the world by far.
It would be similar in size toLinkedIn, but that's not our
data.
However, our investment insourcing was essentially
building the technology to putthis together in the
partnerships so that we have away for us to identify
(40:17):
individuals that you're latergoing to go and research further
Got it.
This creates a very interestingdynamic because essentially,
you can power both recruiterswith this.
Speaker 1 (40:30):
That's great, Elijah.
Do you have any remainingquestions before we jump into a
watermelon photo?
Speaker 3 (40:36):
I'm good.
I'm excited to learn about thewatermelon.
Speaker 1 (40:38):
This is actually one of my favorite questions I'm
asking today.
So, everybody, in the episodedescription there's going to be
a link to Anikas' LinkedInprofile.
So when you click on that,you're going to notice that he's
taking a huge bite from awatermelon.
So I just want to know whatthat's all about and the
backstory to that photo.
Speaker 2 (40:54):
Remember how the you know why I picked the photo to
put on LinkedIn.
It was like a joke or something, but this is one I was using
internally.
But the funny story is this youremember about?
Was it a year or so ago whenSVB collapsed?
Oh yeah, I do remember Workable, had a lot of money on SVB and
(41:15):
I was one of the people whorushed to take the money out
successfully.
And but you remember, I'm greekas well, which means about 10
years ago I did the opposite Irushed to take the money out of
greece when greece wascollapsing oh man so I made the
tweet, like guys, I was neverexpecting I'd be sending my
money to greece to be safe,because for 10 years I've lived
(41:35):
in a situation where this wouldbe a ludicrous thing to say
which got retweeted by somethinglike the Minister of Economics
or anything, and there was a bigdiscussion and then articles
were written and some journalistwrote an article about the
topic in which he introduced mein the article like Nikos
Varaitakis, CEO of Workable, aguy who likes watermelons,
(41:59):
literally on the newspaper, likethe equivalent of the Financial
Times.
So officially for the Greekgovernment, I'm the guy who eats
watermelon.
Speaker 1 (42:08):
It's not a very exciting story, but it's a great
story.
Speaker 2 (42:11):
That's just a bit random, but if I told you I
could bring SBB into that, youwouldn't believe me, would you?
Speaker 1 (42:18):
That's awesome.
Yeah, it's definitely.
It's a memorable photo.
I'm not going to forget that,that's for sure.
But yeah, look, this has been avery insightful episode.
We covered a lot of ground thatwe haven't yet in this AI for
Hiring series.
I think this is the fifthepisode we've done thus far.
We probably have at least 15more to go and because just so
you know is thus far, we've hadfour pretty good episodes if you
(42:40):
want to check them out.
But yeah, we're going to have alot more.
We're doing a combination.
We're talking to some categoryleading companies and very small
kind of startup companies aswell that we think are doing
disruptive things in the space.
So it's a good combinationthere of CEOs that are coming
from different levels of scaleand product market fit, so to
(43:00):
speak, or customer base, andessentially different ways of
architecting these types ofsolutions.
So it's a pretty cool mix right, Like we're speaking to you one
day and then we got a CEO of athree-person company coming on
the next day.
So it's a pretty cool seriesthat we're running right now and
I really appreciate you beingwilling to contribute to the
community, to talent acquisitionleaders across several
(43:24):
industries.
We got listeners in essentiallyevery continent, every country.
I know everybody tuning inglobally is really thankful for
you joining us today.
Speaker 2 (43:32):
Thank you, guys.
I really enjoyed theconversation.
I really do think it's a verycool thing you're doing here, so
thank you for including me.
Speaker 1 (43:41):
Yeah, of course, of course.
Hey, look for everybody tuningin.
We've got a lot of greatepisodes coming up here in the
near term and we got I thinkwe're trying to push out
basically like two episodes aweek for this series.
It's going to be a lot of fun.
Thank you so much.
We'll talk to you next time,take care.
Speaker 2 (43:55):
Thank you.
Welcome to the Breakthrough Hiring Show.
I'm your host, james Mackey.
Today, we are joined by myco-host, elijah Elkins.
Elijah, what's up?
How are you doing?
Hey, I'm all good, good, andour guest today is Nikos
Moriartakis.
Nikos is the founder and CEO ofWorkable and he's going to be
here today to talk to us aboutWorkable, telling us more
(00:21):
insights into the primary valueproposition, how they're serving
customers, and we're going tobe dialing in on AI use cases,
specifically about how Workableis thinking about AI and
feedback they're hearing fromcustomers, and what Nikos thinks
about the recruiting technologyspace right now in the
direction we're heading.
So, nikos, thanks so much forjoining us today.
How are you doing?
Speaker 2 (00:43):
I'm doing well.
Thanks for having me.
It's a pleasure to be here.
Speaker 1 (00:47):
Yeah, Nikos, why don't you just start off telling
us a little bit about yourselfand what you're doing over at
Workable right now?
Speaker 2 (00:53):
Workable is an HR software vendor.
Primarily we cater to small andmedium-sized businesses up to
maybe a few thousand employees.
We've been around for over adecade.
Everybody knows us forrecruiting because that's where
we started, although in the lastfew years we've expanded and
now we do all things HR.
But we've served just to giveyou a sense of proportion, over
(01:16):
30,000 customers over the yearsand we've seen 400 million
candidates for jobs, severalmillion hires.
So it's been around for a while.
We can talk more, a little bitabout what's going to be in 400
million candidates for jobs,several million hires.
So it's been around for a while.
We can talk more, a little bitabout what's going to be in the
future and what we're building,what the customers are doing.
But yeah, we're how can I saymid-size software vendor for HR,
mostly in Europe and the UnitedStates.
Speaker 1 (01:37):
Workable?
Did your team start in the HRISside or the applicant tracking
system side?
Where was the initial product?
And then how did you build uponthat and expand?
Speaker 2 (01:47):
Initially it was a straightforward applicant
tracking system.
We started 12 years ago.
Back then, the main propositionof SaaS software is I'm going
to give you better UIs,prioritize the user, give you
better experience and also maketechnology that previously was
(02:07):
available to the big enterpriseaccessible from price and
utility and ease of use tosmaller companies.
So Workable.
Started like that, like manyother companies of that
generation.
Over time, in recruiting, yourealize that one of the most
important things is sourcing isto be able to build a target
pipeline.
So we spent a lot of timebuilding one of the top partners
(02:31):
for Indeed and one of the toppartners for LinkedIn.
We move around millions ofcandidates every year.
We have our own sourcing tool,so we made an investment in that
and over the last three yearswe've expanded into HR
management, employee profiles,documents, time off, that sort
of thing.
Speaker 1 (02:52):
Yeah, I like the strategy and so it sounds like
you.
When did you expand intosourcing?
I'm just curious from atimeline perspective.
Speaker 2 (03:00):
It was around 15, 16, when we put a lot of emphasis
in sourcing.
Oh, that's awesome, and alsothat happens to be the genesis
of all our AI technology thatwe're going to talk later,
because when you go intosourcing, it's one of the places
where this is even earlier and,let's say, less strong
technology than the one we haveright now, where applicable.
Speaker 1 (03:20):
That's great couple.
That's great because the reasonI asked that question is I'm
curious.
My assumption is that yourcompany weathered the storm of
market consolidation a littlebit stronger than some of your
competitors, due to the factthat you built out a product
suite more so proactively, priorto COVID and the market crash
in 2022.
I've spoken with a lot ofrecruiting technology leaders
(03:42):
that are essentially scramblingto develop a product suite to
make their products stickierright, like maybe they were just
only doing sourcing and nowthey're scrambling like, okay,
we really need an ATS, becausecompanies aren't they're not
going to have severaltechnologies in their stack,
particularly if they're SMB andmid, lower, mid-market focused.
And so then I talked to otherslike who are ATS CEOs that are
(04:04):
like, okay, we really want toget into potentially more of the
onboarding HRS side or we wantto get more into the sourcing
side, and they're essentiallyjust trying to make decisions
based around, likeactively, tryto expand your product suite,
and probably something that Iwould again, this is a total
assumption.
I don't know, but I would say,compared to other recruiting
(04:31):
tech companies, probably enabledyou to have a little bit more
consistency and minimize churnas much as possible, which, of
course, I'm sure for everybodyit's still increased, but
weathered the storm a little bitmore than some of the other
CEOs that I've spoken with overthe past few years.
Speaker 2 (04:46):
Look, you touched on something a little bit more
fundamental about how the marketmoves.
I think we are at the end of acycle of innovation, so usually
you have a decade.
There was a big decade for B2BSaaS software.
It became easier and easier tomake it.
Players played well with eachother, there was a lot of
financing, there was a lot ofappetite in the market.
(05:08):
The market was going up,companies were willing to spend
for point solution, for a smalledge on competitiveness and
there was a lot to be gainedthere and this resulted in a
proliferation of a lot ofsoftware and a lot of software
for very, very specific usecases, from point solution down
(05:31):
to what would be features of apoint solution.
Now, obviously, this clears upat some point.
So naturally, the market isgoing to consolidate.
I think it's driven more bothfrom the industry, as you said.
For each one of those companiesit's a stronger play to be able
to expand into more things, tosell more things, to grow their
footprint on the customer, tomaybe consolidate through M&A
(05:52):
and wicked by the financing withothers and, on the other hand,
the customers as well.
I don't know about you guys,but I hear the story very often.
I am the new HR manager.
Previously, the company wasgrowing fast and somebody bought
expensive software foreverything.
Now cleaning up this mess, Iwant an affordable solution that
(06:13):
does everything.
I bet it's happening everywhereit is.
Now, in our case specifically, Ithink we got a bit lucky.
We were one of the first tocome into this space, so we
built some scale.
We already had numbers that wewere a profitable company with
(06:33):
strong economics.
So, even though the 22 and 23were completely terrible years,
it wasn't a huge setback for us.
And we were lucky as wellbecause, for our own reasons, we
ended up becoming profitableright at the year of COVID,
before we knew that this wasgoing to happen.
(06:54):
So we didn't have to retraceour steps like many companies
had to do during these years,and we said since we're not
going to sell a lot over thenext two or three years, we
might as well make a new productto improve our retention.
Speaker 1 (07:07):
That's pretty smart.
Yeah, lucky or not, I'm reallyhappy for you guys.
It's a strategy that seems toit just makes sense.
It makes a lot of sense andyour team was able to pull it
off, it seems, a few yearsbefore a lot of recruiting tech
companies.
I know recruiting techcompanies are just rolling out
additional products.
In the past year or so theyprobably already churned a lot
(07:28):
of customers that they might'vebeen able to hold onto if they
had done this proactively.
Speaker 2 (07:34):
I think through M&A or product expansion, you're
going to see this market startconsolidating a little bit over
the next few years.
Speaker 1 (07:41):
Yeah, for sure, for sure, Elijah, you have any
follow-up thoughts on that topicfor?
Speaker 3 (07:45):
years.
Yeah, for sure, for sure,elijah, you have any follow-up
thoughts on that topic?
I've actually implementedWorkable a couple of times at a
few different companies.
I didn't mention that to Nikosin the beginning, but so very
familiar with the product.
I haven't used it for a coupleof years but it was super
interesting to see you all hadsourcing before anyone else did,
before the other ATSs, beforeAshby came out with kind of
(08:06):
their sourcing tool with thatkind of all-in-one workable was
had that TA product suite before.
I'm curious you don't see toomany companies right, start with
the ATS, have a very robustsolution and then build out an
HRIS or an HRMS system.
You've seen HiBob BambooHRRippling right try to add like
(08:27):
recruiting modules, like anATS-like functionality to their
HRIS.
What does that kind of looklike for you?
Is the ATS almost a standaloneproduct side by side and you
switch from one user experienceinto the other, or is it
holistically integrated to oneor just yeah?
How does that kind of look tobuild out an HRIS at workable?
Speaker 2 (08:47):
that's a very good question.
It's a different problem whichdirection we go to, and a lot
can be said about this, but letme try to focus on a couple of
interesting bits.
I think what we're doingcommercially is more difficult,
because you may have a bunch ofcustomers who are even excited
(09:08):
about your product but they'renot the buyers of the other
product.
So you're typically not.
You may be talking to the HRdepartment but you're not
talking to the same buyer.
So essentially, commercially,you need to rebuild a lot of the
relationships and marketing andbrand that you need to build.
And it's also usually therecruiting is a sub-department
(09:31):
of whoever buys the HRIS.
So nobody's going to buysoftware for their boss or for
some of the people who havesomething very separate.
So expanding in that manner isnot tremendously easy.
On the other hand, you haveanother force into it that most
companies, as I said, wouldrealize.
It's better to have onesoftware where everything from
(09:53):
the org chart down to theterminations, to the backfields,
to the acquisitions, where itunderstands my people, it
understands who performs well.
Now, with AI, you can believethat understanding performance
is going to link to betterrecruitment.
So when customers go for thisbecomes a question what is
better to buy if you assume,let's say for simplicity, that
(10:16):
the customer says workables HRISis the newest weakest and
bamboos recruiting is the newestweakest, which one do you hurt
more by accepting a lower end?
Speaker 3 (10:27):
solution.
Speaker 2 (10:28):
The argument there, and my hope is that recruiting
is the one that is a performanceapplication.
It does make a big differencein results.
The other one is a databasewith a good UI, let's be honest.
So would you rather have morecandidates or a different
signature?
The signatures are the same.
Speaker 1 (10:50):
When it comes to incorporating AI at Workable,
I'm curious to see howaggressive Workable is with
implementing AI functionality.
My assumption I think youtouched on this is are you
starting more on the sourcingside?
But I'd love to learn moreabout that.
But I think some of the moreestablished players I've seen
are moving maybe a little bitslower, not necessarily because
(11:14):
they have to, but just becausethey have a full product suite.
They're taking their time to bea little bit more thoughtful
about what they're rolling out.
I'm curious are you pushing AIaggressively or are you just
trying to roll it out in verysmall little phases that you're
more confident, or high impact?
What's your kind of AI strategy?
How do you think aboutincorporating AI into a product
(11:39):
or pushing out a feature or likewhere to start?
I'm just curious about, like,your thought process and then
that leading into where did youactually start?
Can you walk us through yourprocess?
Speaker 2 (11:50):
The short answer is that we are mega aggressive with
AI.
We have seen fantastic results.
We already have a ton of thingsin production available to all
our users, and thousands ofusers daily, of many different
things we do with AI.
We're extremely enthusiasticabout the results.
Yes, it's a new technology.
(12:12):
Yes, harnessing it and managingall this is not as simple as it
sounds, but I think everythingis going to move in that
direction and I thinkapplications are going to change
in shape as well.
This is going to change UIs.
It's going to change a lot ofthings, and there's absolutely
(12:32):
no way we're waiting on thisgame.
We want to take every problemthat there is in there and try
to crack it with AI dooms andcrack the code of how it's going
to be represented, interactedwith, et cetera, in the future.
Let me give you a sense ofwhere we are right now.
So today in Workable, we'llwrite your job descriptions with
(12:54):
AI, and already 30% of the jobscreated in Workable today are
written by AI.
This is not a small.
This is a big use case.
We publish 700,000 jobs a yearthrough our customers and a
third of them are already donewith AI.
We have salary calculators basedon our data.
(13:16):
Ai helps you identify whatsalary you need to do.
It creates interview kits, itcreates scorecards, it creates
video interviews for every job.
It functions like a juniorsourcer and it's going to go and
find you 50 passive candidateslike a recruiter would, at least
(13:37):
as a target who are matchingwith the job.
It's going to write the emails,but highly personalized, based
on their experience andeverything that he knows on the
internet about them, to recruitthem.
Those emails have twice theresponse rate of personalized
emails from recruiters andalmost four times the response
rate of templates.
(13:58):
And then, for all the candidatesthat come into the system,
anything that you put in thatyou see on your interface we
scan all of them.
For each job description, wehave an AI that essentially
determines what should be thecriteria for screening for that
job and then it marks every oneof those criteria.
So, basically, on top of everyprofile you see a little
(14:21):
scorecard what are the fivebullets you want and which ones
are green.
So it it does all the readingof the resumes and
interpretation and screening foryou, and all of these things
are live use cases like they'renow on the system.
You can go get a trial accountand use them, and thousands of
people are using them every day.
So there's an endlessdiscussion about where this is
(14:44):
going, how good it is, and allthat good stuff, but it's clear
to us that this is going to be aproductivity bomb.
Speaker 1 (14:53):
Okay, so we covered a lot of ground there too.
I actually had to write somestuff down because I want to
slow down on it.
I wasn't, honestly.
I thought you were just goingto tell me like, oh, we were
going to, we can write emailsfor your sourcing team, like a
lot of companies.
That's where they're at.
Oh, we generate a sourcingmessages for you.
Okay, we talked about, ofcourse, like the AI generated,
jd, which is, honestly, that's.
(15:14):
A lot of companies are doingthat one Interview kits,
scorecards.
Now for the interview kits.
Can you provide us a little bitmore detail on what that
includes?
Speaker 2 (15:26):
I'd like to dial into that a little bit more
Basically.
For an interview kit you arepretty much the best way to
describe it.
You're pretty much a structuredinterview.
Okay, you have some areas thatyou want to touch and for each
one of them you have a fewquestions and next to it, if you
have a scorecard, you also havea way to note how well you felt
(15:47):
that person responded to thatquestion on your notes.
Now, this is a great practice.
The problem is to construct thisset of questions and to have a
comprehensive interview thatanswers the right points and
have imaginative ideas forquestions.
That's hard.
That's what the AI does.
It basically writes that foryou.
So, let's say you were hiring afinance director, it will
(16:09):
create a section about.
It says it sees director.
It will create a section aboutleadership, managing teams, et
cetera, and have some questionsabout this.
But he will also ask questionsabout the specific jobs and if
you do a video interview, hewill create the video interview
questions for you.
So he will prepare the videointerview, what should be asked,
(16:30):
etc.
Which is a slightly shorterform version of a scorecard.
In fact, the thing.
Okay, this is new.
We have a way as well to forthis, to listen in on the
interview and transcribe thenotes.
But the funkiest part thatwe're going to do soon is that
essentially it's going to go tothe scorecards and evaluate the
(16:52):
candidates based on what itheard on the interview.
It's going to write theevaluations for you, so it
drafts it for you, you check it,that it's fine, you make any
corrections and you're done.
And then the other thing thatfor next year is that obviously,
if you're looking at a lot ofpeople, you would like to see an
inference report, somebody likea recruiter telling you we've
looked at these five candidates,this one was stronger.
(17:14):
They're going to write you abit of an essay where it
synthesizes for every area who'sthe strongest and et cetera.
So basically the other thingthat we're making now we have
shown a little video of a demoof it.
We haven't made it public yet.
We have an agent.
He has an email in a Slack.
(17:35):
You talk to it on Slack likeyou would with an employee and
essentially it goes and doesthese things I just said for you
.
So you say schedule the nextones for interviews, move them
along, tell me what was theresult of all the interviews,
start to make the job andbasically you talk to it like
you would an employee and itgoes and accomplishes tasks and
(17:56):
it also comes back to you andsay I got three emails from
those candidates.
One is not interested, I willreject him.
The other two are interested.
I prepared a response for youto review.
Click here and go on.
This is how you should imagineit in the future.
So anyway, I'm getting excitedwith these things and I just go
(18:17):
off on a tangent, so I'm sorry.
Speaker 1 (18:18):
Oh yeah, the tangents are the best part.
Let's just go down like therandom stops where we just go
down these random rabbit holes,which is like the whole point
here.
I think that's again the valueout of this podcast in general
is the hosts, as well as theguests, are experts in the space
, so we can get a little bitmore technical, which is, I
think, what we like to do andwhat our buyers and what
Workable's buyers they want tohear.
(18:39):
They want to understand thedifferentiators, how and even
it's not only just like yourcurrent functionality, it's like
how you think about AI, becausethat's going to inform your
product development and roadmap,right, if they feel like, okay,
this guy, nikos, really has asolid pulse on this stuff.
He's doing some innovativestuff in the space that we're
not hearing from otherwell-known recruiting tech
providers.
Like, maybe we should bet onthis guy and workable versus
(19:00):
some of the other ones.
So it's going down these rabbitholes is actually, I think,
really important.
Yeah, it's fun for us, first ofall, selfishly, but it's also
something that I think peopleneed to hear.
So I guess, like I and Elijahjust cut me off.
Man Like I'm, I have so manyquestions, questions, but I I
like one one thing real fast isnow you're talking about, like
the video interviews you talkedabout.
(19:21):
Essentially, are you guys alsodoing like the ai note taking,
where you're plugging intointerviews and it sounds like
you said you're going to startto put together evaluations of
the transcript.
Speaker 2 (19:31):
so you're doing that, okay.
But the interesting thing isthen when you read the
transcript Okay, so you're doingthat, okay.
But the interesting thing isthen when you read the
transcript and you alsointerpret it and help the
interpretation.
But that is for later.
Speaker 1 (19:44):
So for the interpretation piece like what's
interesting is that we've nowspoken with the CEO of
BrightHire, ceo of Pillar andthen a couple of CEOs of smaller
companies a little lessestablished in the space, and
what was interesting aboutPillar and BrightHire is neither
of them are actually providing.
It's like they're providingsummaries and they're packaging
(20:07):
the data in a consumable way sothat hiring managers can compare
and contrast candidates andthen also evaluate.
Okay, you have three stages ofinterviews, you've covered 75%
of the things you need to, butwe didn't ask this one question
in the screening call.
We need to double back on that.
Or they're providing some kindof have, we asked all the
evaluation criteria, but they'renot necessarily scoring
(20:28):
candidates.
It doesn't sound like theproducts are saying, hey, here
are your best candidates, orthey're not really doing an
evaluation piece or a scoringpiece.
We spoke to a CEO of a smallercompany called Qual and they're,
I think, a very early stagecompany.
They do scoring.
They do scoring.
Now it's different.
(20:49):
They're working with highvolume, light industrial workers
where apparently, he said, theresumes are typically very
incomplete so it's verydifficult to tell who's worth
speaking to, and so the scoringaspect is important because
there's thousands of applicantsand there has to be some kind of
way.
That's not doing a keywordmatch.
That's not effective.
So maybe because Bright IronPillar are more in the corporate
(21:12):
type of space, if you only havefive candidates in the
interview process, do you reallyneed a stack rank or like a
scoring mechanism in place?
I'm curious how you look atthat topic, like when you're
thinking about the note-takingaspect, packaging the analysis
essentially, or summarizing.
Do you see Workable going inthe direction where you're
(21:32):
actually going to be scoringcandidates and saying hey, or do
you see yourself staying awayfrom that?
And how important do you thinkit is?
Why would you do it or not?
What do you think?
Speaker 2 (21:42):
No, because I don't think that's what you need in
recruiting.
I think most people, especiallypeople outside the space,
imagine that recruiting is aprocess that involves a great
degree of judgment and rankingpeople and making some
evaluative assessment on them.
(22:04):
In reality, it's not so much.
It's more of a filteringprocess.
It's a process where you try toarrive at a few people for whom
you have collected theappropriate information and you
have verified that they are anappropriate fit.
Let me put it in simpler wordsIf you give me a short list of
five resumes of people who workthis job, have the right
(22:27):
qualifications and we haveverified a few things around
them, my job is easy.
That's not what I want thesoftware to do.
That's an easy task.
The computing part in recruitingcomes into the fact that you
have to process a lot of thingsfrom many sources and you have
to go through many steps.
It involves many channels, fromjob boards to emails, to this
(22:48):
to that.
So it's more of an informationorganization process, as you
suggested.
We see, as you said before,many vendors say I'm going to
organize the information for you, I'm going to present it to you
, transform it, process it.
Take something that was said inan interview, write it in words
, put it in the right box so youcan see it next to the right
(23:09):
things and make a decision.
The decision is neither themost important thing for the
software to do, and I think it'sthe thing that the software is
least qualified to do, becauseit's the most idiosyncratic as
well.
Think about it Honestly at theend of the day, people do have
companies, departments, teams,whatever it is.
(23:30):
They have different view ofwhat is important to them.
The point is to get them to thepoint where they have a
meaningful choice between a fewvery qualified candidates for
which they have done their duediligence, so to speak.
They've got the rightinformation.
When you hire an executiverecruiter and you pay top dollar
(23:52):
for that job, what do you get?
You get all the resumes, youget all the information, but you
also get some organized dataand comparative things around
them and you get a little bit ofa narrative.
Somebody, essentially, hasprepared a dossier with you and
has checked everything, and heknows that you could hire any
one of those people.
(24:12):
Just pick the one you like.
That's what the software shoulddo.
The other thing about scoringand ranking it's also that it's
touchy.
You can really go wrong there.
Do you want a world where anyvendor can make a tweak on the
algorithm and that changes whogets hired?
(24:32):
We don't want to get into that.
I don't know if I'm willing tobelieve that sometime in the
future AI might be better andmore objective than humans,
because humans are flawed forthings like that.
Maybe sometime in AI is goingto be better, but not today, not
next year.
Speaker 1 (24:55):
What about top of funnel?
So okay, not hiring decisionscoring and I don't mean scoring
necessarily like 98%, I justmeant okay, of the criteria that
you're looking for.
There's a group of folks thatseem to be more aligned with
that.
Whatever like loose, right?
Okay, so not.
Maybe not at the end of theprocess.
It's more packaging, the data,the narrative, as you put it,
organizing, I think, the slackaspect of what you talked about.
(25:16):
Hey, reach out to these nextfolks.
That's awesome.
I'm really excited to see thatproduct.
That's going to be so cool.
But what about top of funnel?
Right, so you get 1,000applicants?
Do you see value in any kind ofautomated screenings?
So AI can help package, saying,okay, out of these 1, these
thousand people, you'reessentially allowing them to
(25:38):
engage in a bit of the screeningprocess upfront so that you can
have a better way ofidentifying a short list than
just do a resume keywordmatching situation that, either
done by a person or anapplication, is not necessarily
the best way to figure out whoto actually jump on a call with,
like with a person.
So do you see any kind ofscoring benefit on top of funnel
(26:01):
?
Or still you're just you don'tlike it?
Curious there.
Speaker 2 (26:04):
Even on the top of the funnel, where indeed there
is more motivation, because youreally need to do a first filter
and, to be honest with you also, we're not the government.
We're not giving peoplebenefits to the extent that it's
reasonably objective, even ifthere was someone who was good
and you didn't manage to capturethem, it's survivable and
humans do it too.
However, I think it isextremely hard to create a
(26:31):
scoring mechanism that isn'tgoing to go wrong.
It's such a complex problemlike.
Even if you have five criteria,we all know they do not all
have the same weight.
The job description doesn'ttell it either.
It's just a document.
It's not in your head, so Idon't.
I would be extremely reluctantto make a system.
We could make a system likethat with Merka, but it would be
(26:53):
good many of the times.
It could probably beat some notvery good recruiters.
That's not what you need, whatyou really need.
In screening.
You have a big task.
You have to go through a lot ofresumes and you have to
impartially, objectively andconsistently evaluate the facts
(27:15):
on the resumes versus a list ofthings you're looking for.
Nobody really does this.
It would take you a few minutesper resume.
Nobody really does this.
In fact, most people, actualrecruiters what they do they
operate more like keywordmatches.
They look for the resume for afew things that seem to light a
bell on them and then they movethem forward.
And let me tell you thebusiness reality of the thing.
(27:38):
A few more resumes come a fewdays later.
Do you think they get the sameattention from the same person
with the same criteria?
So there's a very inconsistentprocess from getting to the big
amount of information that's abig chunk, which is all the
resumes to the summary of that,which is all these people on the
five things we need will do thesit.
That is what AI can do for you,and then AI doesn't need to be
(28:02):
involved in the decision becauseit has condensed the
information, does the legwork somuch for you that now it's easy
for you to just look at themand apply your own criteria on
top of that.
Speaker 1 (28:13):
It's like a data organization presentation.
Speaker 2 (28:17):
I wouldn't do scoring , because scoring suggests that
you take all that information toa formula and say this guy is a
10, is a 20, is a 50.
That is what does it mean.
It sounds silly even when yousay it.
What does this mean?
But what it can really do is toshow you a bunch of cards of
people with some informationthat makes you easy to make a
(28:40):
decision.
Let's think of it realistically.
You see, someone who meets allfive criteria, move them on,
someone in between?
Okay, I can still look andverify, but it really makes it
easy to pick up things that arecompletely off or completely on.
Which is the majority?
(29:02):
Okay, when you get the pipeline, the majority.
If you take the ones that arebang on and the ones that are
white in the sky, apply here youhave 80% of it.
Speaker 1 (29:11):
Yeah, that's okay, that's super helpful.
Thank you, elijah.
What do you got, man?
What questions you got in yourhead?
Speaker 3 (29:17):
Yeah, two quick questions.
One's a comment.
Thank you, elijah.
What do you got, man?
What questions you got in yourhead?
Yeah, two quick questions.
One's a comment, the other is aquestion.
Have you guys considered or areyou working on anything that
analyzes all of the scorecardsand the feedback and provides
almost like a state of the roleand kind of trends that seem to
be things going well withcandidates being passed through,
(29:37):
or things not going well,because I've never seen that,
but that's often right.
What a more strategic kind ofrecruiter is trying to figure
out is what are these trendswith the candidates being passed
through?
Why are candidates gettingrejected?
And you don't always have theright rejection reason
categorized appropriately in theATS.
It may actually be within dataor the scorecard notes or
(30:01):
something.
Does that make sense?
Speaker 2 (30:04):
Yes, but not with AI.
Workable allows you to do this,both to classify the reasons of
rejection and obviously whenyou get the scorecard.
There is a little bit more howcan I say?
Informal enforcement.
It's very hard for someone whomeets all five criteria to say
it wasn't good enough on thefirst screening.
(30:25):
So it keeps you honest a littlebit because it surfaces more
information more deliberately infront of you, but automatically
, no, we don't do it.
Now, the other thing you saidto try to even create a profile
for the role.
I haven't thought of it thisway.
What I think, what I would likeas a user, I would like the AI
(30:46):
to get to the point where, inone way or another, it tells me
remove this requirement.
It's stupid, you're notselecting for it.
You said five years ofexperience and you didn't honor
it.
Speaker 1 (30:58):
stop I wish I could talk to my customers that way as
a like secure vision bycompanies rpo provider.
I wish I could get on the phonewith a demanding ceo of a
growth stage sass company and belike remove this requirement.
It's stupid, I end up doing it,but it takes a lot more tact
and time and data.
To summarize, but if I couldhave a software do it for me,
(31:19):
that'd be great.
I could say, hey, it's allright here.
Speaker 2 (31:23):
What we are hoping at Workable and I'm not going to
say that we really do this inany meaningful way, okay, only
indirectly.
Is that eventually Workablewhen it writes the job
description which, as you said,to get a GPT or something
similar to write?
It is not very impressive onits own.
It's to the point where itstarts writing the job
(31:44):
descriptions for better successfor you knowing that stuff.
But right now you will seeapplications using AI technology
let's say, black boxes for themoment, using this new
technology, plugging it intospecific places and improving a
part of the experience and thenmaking it informed by more
(32:04):
information.
And when we collapse the dam ofinformation and start sharing
between for your account, whereone can inform its decisions
from things that happenelsewhere in the application
that you might see the reallybig steps change, where this
thing makes things better foryou because it has also observed
(32:26):
the way you work.
We're not there yet, but wewill be there in a few years.
Speaker 3 (32:34):
Is that part of the building?
The HRIS side of things isreally so you get that whole
picture of, like talentmanagement, who's being promoted
, who's at the company, theirperformance, and that can
actually feed back into therecruiting side of things.
Speaker 2 (32:50):
Yeah, let me tell you futuristic features I want.
When somebody quits to have aclone button and this thing
understands this employee, theirrole, who they work with.
It probably already knows theirpersonality tests etc.
And then goes out to source andrecruit people against that and
(33:11):
even tells you how theirpersonalities will fit Remove
the personality part, whichsounds a little bit dodgy.
Something that understands yourcompany, your needs, something
that gets your hiring plan andknows how much time it's going
to need to fulfill thosepositions, because it takes us I
don't know three months to geta batch of HDRs.
So I want all my orgs to beable to do my future planning
(33:33):
and my acquisition planning andmy succession planning.
And this thing understands thepeople.
I want a referee on theperformance review, somebody who
has gone into salesforce and toapplication and has read things
about the employee and knowsthe one-on-ones and tells me and
helps me do the performanceevaluation and compares people
(33:54):
who's the best in the departmentof this or that.
There's so many things you cando.
This is futuristic.
I'm not promising somethinglike that now, but it is totally
within the realm of thepossible and I think the most
interesting thing with AI is notgoing to be that it's going to
give like one absolutely wow upthat stuff.
Like that are going to startsounding.
Ai is going to get intoeverywhere and eventually smart
(34:17):
companies are going to startsounding.
Ai is going to get intoeverywhere and eventually smart
companies are going to createinteresting UIs that do these
things.
Imagine we're in 1997 and say Iwould like a place where it has
all the videos and all themovies in the world and I can
watch them by connecting to theinternet.
It would happen.
Speaker 1 (34:36):
Similar things would happen with that when we were
talking about some of theinitial ai applications, there
was a follow-up question I hadon sourcing.
But before I switch over tothat topic, elijah, do you have
any other questions you want todial in on here?
Speaker 3 (34:48):
I'm just curious quickly, nikos, how do you feel
like workable and other but?
what I wouldn't call legacy ATSsright, because it was only 10
years ago.
How do you feel like you'llcompete and what the dynamic
will look like?
Competing against more like AInative maybe you could say
(35:09):
recruitment software companiesor HR software companies that
are coming out of places like YCombinator today, where they're
starting with Gen AI day one andarchitecting everything around
that?
Do you think companies likeWorkable still have the
advantage because of theexpertise and knowledge and team
size and customers and all ofthat, or that they actually may
(35:31):
have an advantage in some ways?
I'm just curious.
Speaker 2 (35:36):
Look, any new company has the advantage that they're
starting from a blank slate.
Okay, this in some years, mostyears is not an advantage, it's
a liability when there's a newtechnology coming into place and
everybody needs to adapt.
What did I say in the beginning?
(35:56):
This is going to change UIs andhow we think about how this
happens and all that stuff.
Obviously, if there's a newoptimal state, there's a new
best design for that thing witha new technology, the one who
starts from a black slate has achance of getting there faster
than the one who has to redesignthings there faster than the
(36:16):
one who has to redesign things.
So the question becomes where isthe change going to happen?
Right now there is.
I don't want to strawman it,but there is a little bit of a
how can I say?
Simplistic view that, oh, sinceAI can do everything, we can
(36:37):
write the code.
We can do everything with AI.
Basically, we replace a lot ofthe business logic,
functionality, workflows,processes, right, yellow that
the software had with just an AImaking a decision or making it.
It's not good enough to getthat point.
You can't just get an AI andsay, oh, you be my recruiter and
now we have no UI.
I just talk to you and you doeverything.
(36:58):
We might get there one day, butI'm not even sure that would be
the ideal interface for a bigcompany.
But I do not fear that somebodyis going to use AI to replace
what a big business softwarelike Workable does.
I'm not afraid of that at all.
I think the technology is goingto come and slot in.
(37:18):
Just because you made a newplane, you can't throw away your
airports.
Somebody's going to post thisjoke to LinkedIn.
There's so many things thatneed to happen.
What I think will be disruptiveis the second wave of those
companies that come in withdifferent premises.
Sooner or later we will startto see an emerging pattern on
(37:40):
how new software is going to beorganized and new UIs, somebody
coming with a good intuition andvision about how that would
apply to the particular domain,like we did 12 years ago A
different design of how theapplication should be.
And they use AI technology toget there much faster.
(38:02):
And because they built thiswith the assumption it's going
to be done with AI, they are notincubates from the requirements
of our users, who would not bewilling to put up with such a
dramatic change, I'm guessingI'll tell you in five years.
All right, sounds good, butwe're not willing to put up with
such a dramatic change.
Speaker 3 (38:20):
I'm guessing I'll tell you in five years.
Speaker 2 (38:21):
All right, sounds good, but we're both heading in
the same direction, and that'sthe important thing.
Speaker 1 (38:25):
Good.
So I have one question aboutsourcing, and then I was hoping
I could ask you about yourLinkedIn profile photo.
We're eating the watermelon.
I want to make sure we haveplenty of time for that.
But when it comes to thesourcing functionality, we had
plenty of time for that.
But when it comes to thesourcing functionality, you're
saying it's like the AI can goout and pull candidates right,
like I think you said that right, the AI could say, hey, here's
some great candidates for you toconsider, right, okay.
(38:45):
So if that's the case, thenlike, where does it go?
Where is it pulling this datafrom finding the candidates?
How's it doing that?
Speaker 2 (38:54):
Earlier, I mentioned that for a few years we spent a
lot of time developing sourcingtechnology.
Essentially, what we've builtis through a network of
partnerships with job sites,with data providers, etc.
We've built a pretty decentdatabase for looking for talent.
It's not as big as LinkedIn.
(39:17):
It's not as accurate.
It's basically a composition ofmany sources.
However, for the purposes of anAI identifying people that
would be suitable or a goodmatch for the job, it's
perfectly good because it's muchyou're going for.
So, essentially, we have ourown proprietary database that
(39:37):
we've built over the years.
Just to be clear, this doesn'tinvolve our customers, right?
Yeah, obviously, I'm assumingyou guys understand that.
There's no way this could bethe same thing.
Speaker 3 (39:49):
Yeah, yeah, yeah, different data.
Data is completely separate.
Speaker 2 (39:54):
Obviously otherwise, if we could, if the data of our
customers was our own which itisn't it would be one of the
biggest in the world by far.
It would be similar in size toLinkedIn, but that's not our
data.
However, our investment insourcing was essentially
building the technology to putthis together in the
partnerships so that we have away for us to identify
(40:17):
individuals that you're latergoing to go and research further
Got it.
This creates a very interestingdynamic because essentially,
you can power both recruiterswith this.
Speaker 1 (40:30):
That's great, Elijah.
Do you have any remainingquestions before we jump into a
watermelon photo?
Speaker 3 (40:36):
I'm good.
I'm excited to learn about thewatermelon.
Speaker 1 (40:38):
This is actually one of my favorite questions I'm
asking today.
So, everybody, in the episodedescription there's going to be
a link to Anikas' LinkedInprofile.
So when you click on that,you're going to notice that he's
taking a huge bite from awatermelon.
So I just want to know whatthat's all about and the
backstory to that photo.
Speaker 2 (40:54):
Remember how the you know why I picked the photo to
put on LinkedIn.
It was like a joke or something, but this is one I was using
internally.
But the funny story is this youremember about?
Was it a year or so ago whenSVB collapsed?
Oh yeah, I do remember Workable, had a lot of money on SVB and
(41:15):
I was one of the people whorushed to take the money out
successfully.
And but you remember, I'm greekas well, which means about 10
years ago I did the opposite Irushed to take the money out of
greece when greece wascollapsing oh man so I made the
tweet, like guys, I was neverexpecting I'd be sending my
money to greece to be safe,because for 10 years I've lived
(41:35):
in a situation where this wouldbe a ludicrous thing to say
which got retweeted by somethinglike the Minister of Economics
or anything, and there was a bigdiscussion and then articles
were written and some journalistwrote an article about the
topic in which he introduced mein the article like Nikos
Varaitakis, CEO of Workable, aguy who likes watermelons,
(41:59):
literally on the newspaper, likethe equivalent of the Financial
Times.
So officially for the Greekgovernment, I'm the guy who eats
watermelon.
Speaker 1 (42:08):
It's not a very exciting story, but it's a great
story.
Speaker 2 (42:11):
That's just a bit random, but if I told you I
could bring SBB into that, youwouldn't believe me, would you?
Speaker 1 (42:18):
That's awesome.
Yeah, it's definitely.
It's a memorable photo.
I'm not going to forget that,that's for sure.
But yeah, look, this has been avery insightful episode.
We covered a lot of ground thatwe haven't yet in this AI for
Hiring series.
I think this is the fifthepisode we've done thus far.
We probably have at least 15more to go and because just so
you know is thus far, we've hadfour pretty good episodes if you
(42:40):
want to check them out.
But yeah, we're going to have alot more.
We're doing a combination.
We're talking to some categoryleading companies and very small
kind of startup companies aswell that we think are doing
disruptive things in the space.
So it's a good combinationthere of CEOs that are coming
from different levels of scaleand product market fit, so to
(43:00):
speak, or customer base, andessentially different ways of
architecting these types ofsolutions.
So it's a pretty cool mix right, Like we're speaking to you one
day and then we got a CEO of athree-person company coming on
the next day.
So it's a pretty cool seriesthat we're running right now and
I really appreciate you beingwilling to contribute to the
community, to talent acquisitionleaders across several
(43:24):
industries.
We got listeners in essentiallyevery continent, every country.
I know everybody tuning inglobally is really thankful for
you joining us today.
Speaker 2 (43:32):
Thank you, guys.
I really enjoyed theconversation.
I really do think it's a verycool thing you're doing here, so
thank you for including me.
Speaker 1 (43:41):
Yeah, of course, of course.
Hey, look for everybody tuningin.
We've got a lot of greatepisodes coming up here in the
near term and we got I thinkwe're trying to push out
basically like two episodes aweek for this series.
It's going to be a lot of fun.
Thank you so much.
We'll talk to you next time,take care.
Speaker 2 (43:55):
Thank you.
Popular Podcasts
24/7 News: The Latest
The latest news in 4 minutes updated every hour, every day.
Stuff You Should Know
If you've ever wanted to know about champagne, satanism, the Stonewall Uprising, chaos theory, LSD, El Nino, true crime and Rosa Parks, then look no further. Josh and Chuck have you covered.
Dateline NBC
Current and classic episodes, featuring compelling true-crime mysteries, powerful documentaries and in-depth investigations. Follow now to get the latest episodes of Dateline NBC completely free, or subscribe to Dateline Premium for ad-free listening and exclusive bonus content: DatelinePremium.com